MTI Report 04-03
The
Development Strategies, Location Decisions, and Travel Characteristics along a
New Rail Line in the
April 2005
Dr. Hollie
Lund
Dr. Richard W. Willson
a publication of the
Mineta Transportation Institute
ACKNOWLEDGMENTS
This study was a collaborative effort
by a team of researchers and university students. The authors would like to
thank the following persons for their efforts and suggestions: Researcher
Daniel Evans,
Additional thanks to MTI staff, including Research Director Trixie Johnson, Research and Publications Assistant Sonya Cardenas, Graphic Designer Shun Nelson, Webmaster Barney Murray, and Editorial Associates Irene Rush and Catherine Frazier for editing and publishing assistance.
Table of Contents
The Gold Line and Its Station Areas 15
Overview of Residential Surveys 29
Demographic Characteristics 30
Residential Location Choices 39
Comparison to Prior Commute Behavior 48
Analyses of Variance in Transit Ridership 49
Overview of Interview Findings 57
Property Managers and Marketing 78
Conclusion and Policy Recommendations 83
List of Figures
Map of Pasadena Gold Line and its rail stations 15
Gender of survey respondents 31
Race/ethnicity of survey respondents 31
Number of persons living in respondent's household 32
Availability of vehicles in respondent's household 33
Number of years that respondent has lived at current residence 34
Occupation of respondent (self-reported) 34
Respondents' approximate household income (after taxes) 35
Share of respondents who work (or attend school) away from home 36
Transportation options available at TOD residents' place of work or school 37
Respondents' perceptions of the local transportation environment 38
Respondents' attitudes toward three dimensions of automobile travel 39
Top three factors considered when moving to current TOD residence 40
Housing and transportation-related location factors by station area(s) 41
Transit commuting frequencies for all surveyed TOD residents combined 42
Primary modes of travel for reported main trips (all TODs) 43
Primary modes of travel by station area(s) 44
Parking price and availability at destination (vehicle trips only, n=304) 45
Primary trip purposes for reported main trips (all TODs) 46
List of Tables
Summary of data collection efforts by station area 10
Key demographic, social, economic and housing
characteristics for
one-third mile radius around station area 25
Gold Line survey response rates by station area(s) 30
Comparison of prior and current commute modes for surveyed TOD residents 48
ANOVA in transit ridership: limited household vehicle availability 49
ANOVA in transit ridership: office or professional occupations 50
ANOVA in transit ridership: annual household income of $15,000 or less 50
ANOVA in transit ridership: respondents over the age of 65 51
ANOVA in transit ridership: recently moved into station area 52
ANOVA in transit ridership: moved into station area for transit accessibility 53
ANOVA in transit ridership: moved into station area for highway access 53
ANOVA in transit ridership: moved into station area for cost of housing 54
ANOVA in transit ridership: moved into
station area for type or
quality of housing 54
ANOVA in transit ridership: commute vs. nonwork travel 55
ANOVA in transit ridership: trips that involve additional stops 55
ANOVA in transit ridership: Los Angeles trip destinations 56
ANOVA in transit ridership: Pasadena trip destinations 56
Summary of developer interviews 60
General perceptions of TOD in the Gold Line corridor 63
Effects on land acquisition, market potential, financing, design, and parking 67
City approvals and future valuations 74
Executive Summary
Overview
A study recently prepared for Caltrans
(Travel Characteristics of Transit-Oriented Development in
To take a more comprehensive approach to the study and
understanding of rail station development within the
Research Context
The Pasadena Gold Line extends from
Union Station, the Southern California rail and bus transit hub located at the
northern edge of downtown Los Angeles, to Sierra Madre Villa on the east side
of Pasadena. The Gold Line intersects the cities of Los Angeles, South
Pasadena, and Pasadena and travels through a wide range of environments, from
L.A.'s Chinatown, to upscale neighborhoods of South Pasadena and the Old Town
area of Pasadena, to suburban employment centers on the east end. The total
travel time (according to the Los Angeles County Metropolitan Transit Agency
[LACMTA], the operating transit agency) is 36 minutes. The line has 13
stations, including the terminal points at Union Station and Sierra Madre
Villa. Because this study concentrates on suburban and infill development, as
detailed below, data collection is limited to TODs outside the central
The Gold Line carried 17,018 weekday riders as of July 2004, approximately one-half the forecasted ridership. The total ridership of the Long Beach Blue Line is significantly higher, but weekend ridership on both the Gold Line and the Blue Line is relatively strong.
Study
Purposes
The first purpose of this study is to examine the early travel behavior outcomes of Gold Line TODs and identify TOD design and policy features that contribute to success. The project methodology is consistent with that used in Travel Characteristics of Transit-Oriented Development in California to allow comparisons to results elsewhere in the state and to complete the creation of a baseline for future Gold Line or statewide studies. The study methodology is also similar to studies done on this subject in the early 1990s. 3
The second purpose of this study is to understand the factors behind the TOD boom in the Gold Line corridor. Unlike the previous studies, this research includes interviews with key developers and property managers operating in Gold Line station areas. In these surveys, they discuss their understandings of the influence of rail transit accessibility on development, to identify both encouragements and impediments to TOD.
The project has the following policy implications:
It provides an information base that can enhance station-area and access planning for the Gold Line. This can help cities decide on the development densities and land-use combinations that are most appropriate for specific station areas.
It informs planning for extensions to
the Gold Line, one that will proceed to East Los Angeles (anticipated to open
in 2009), and the other that is proposed to proceed east, well into the
It assists developers in assessing opportunities and designing sites.
By comparing these data to
information about other rail lines in
Literature Review
In 1992, Robert Cervero conducted a statewide study of travel
behavior among Californians who live or work near rail stations. 4 Because
the BART system is the most developed rail system in the state, the research
focused primarily on the Bay Area, but some sites from Southern California and
Both studies come to three similar conclusions:
Ridership is (on average) around five times higher among people who live or work near rail stations compared to surrounding areas.
The likelihood of a person's using transit is influenced by characteristics not only of the trip origin, such as density and distance to station, but also at the trip destination, such as parking costs and bus connectivity.
Distance to the station is more critical than land-use mix or local design elements in predicting transit use--in other words, a person who lives or works close to a rail station and wants to use transit is not likely to be deterred by an unfriendly pedestrian environment.
The follow-up study also concluded that although overall transit ridership had not increased significantly, BART was finally achieving its TOD ridership expectations. Both of these studies help us to understand residents' and office workers' reactions to rail access and provide a strong basis for future longitudinal studies of travel behavior along California's rail lines, but neither provided insight into station-area development processes nor examined a new rail line. These are the primary research gaps being served by this study of Gold Line station areas.
TOD Implementation
Two other
The nonridership advantages of TOD can counteract some of these
barriers if they coincide with local goals. A rail station's ability to
leverage development opportunities, for example, can help to revitalize
distressed areas or downtowns and increase the use of vacant areas. 8 TODs
also provide improved opportunities for housing, particularly in areas
suffering from a housing crisis. 9 In
TODs and Travel Behavior
Past research on travel characteristics within TODs strongly
indicates a link between physical factors, such as density and transit
accessibility, and increased transit ridership. 11 TOD implementation,
however, does not always result in increased transit ridership, and the roles
of demographics, transportation investments, and other policies and development
strategies need to be considered. For example, the Blue Line has the highest
ridership of
Transportation infrastructures can also inhibit ridership, in spite of investments in TOD and rail. For example, rail networks are often oriented toward central business districts, despite the decentralization of regions and a loss of jobs in downtown areas. 13 This can reduce the use of rail for commuting purposes. A study in 2002 attributes lower-than-expected modal shifts (from automobiles to transit) to parallel developments in automobile and transit infrastructure. This study concludes that transit ridership will not occur at the desired levels as long as similar investments are being made in automobile infrastructure. 14 This observation has been noted by several other studies. 15 Parking supply and pricing--especially in suburban areas, where parking is often oversupplied and constituents are less supportive of priced or reduced parking to facilitate TOD 16 --is also a significant deterrent to transit ridership among residents and users of TOD. 17
Finally, it is becoming apparent that household location choice
and residents' attitudes toward transit access play an important role in TOD
travel behavior. A longitudinal travel diary study of residents in the
By taking a combined quantitative-qualitative case-study approach
to understanding TOD, this research not only expands on our current
understanding of TOD implementation, travel behavior, and residential location,
but also explores and identifies connections among these factors and raises new
questions for future study. This research also expands on the existing research
by specifically targeting TOD along a new rail line--research that will be
increasingly important as
Research Methodology
This report presents findings from three related but distinct
data collection efforts. First, residents living within one-third of a mile of
a Gold Line rail station were surveyed regarding travel behavior and
residential location. The sampling unit for these surveys is the multi-unit
building; sites were limited to infill and suburban locations outside of
downtown
The second data collection effort consisted of qualitative interviews with developers and property managers of residential TOD buildings in Gold Line station areas. These buildings do not necessarily overlap with the residential survey sites because many were not yet occupied and some were not completed at the time this research was conducted. Interviewees were identified using a snowball sample technique, in which public sector informants provided contacts in the Gold Line development community.
Finally, data on each station area were compiled through secondary data sources and site evaluations. Quantitative assessments of the station area population and housing characteristics were compiled from 2000 U.S. Census data, and the local pedestrian environment was assessed quantitatively through site evaluations (see See Appendix C: Pedestrian Evaluation WorksheetSee Appendix C: Pedestrian Evaluation WorksheetAppendix C). These data were used for comparison purposes and to improve our understanding of the household survey and interview findings.
Research Objectives
There were three primary objectives for this project:
To understand property developers' and managers' perceptions of
and responses to development opportunities within new
To evaluate station-area residents' responses to the new rail access in terms of travel behavior and location decisions
To provide a more complete data set for future studies of how travel choices, location decisions, and property management decisions within station areas change over time.
Selection of Survey Sites
As noted above, interviewee subjects were identified through a
snowball sample technique, beginning with public sector informants. Identifying
potential survey sites required extensive legwork. The first task in selecting
sites for study was to create a list of multi-unit buildings located within
one-third mile of the Gold Line stations. To increase our range of potential
sites and to recognize the significant variation in development patterns across
the station areas, buildings of all sizes were considered. (This is different
from the previous studies, where buildings were limited to 50 units or more in
order to make the most efficient use of resources.) Students in a senior
Community Planning Studio class in the Department of Urban and Regional
Planning initiated this task, conducting field reconnaissance on all Gold Line
stations. After the students listed potential sites, a research assistant
refined and narrowed the list based on additional criteria. It was important
that the selected sites represent a range of housing affordability; sites in
the cities of
All sites were located within walking distance (one-third of a mile) of a rail station, except in the case of Sierra Madre Villa (the terminal station), where we had to extend out to a 1-mile radius to find suitable multi-unit buildings.
Household
Surveys
The quantitative assessment of travel behaviors and location decisions was conducted through self-administered questionnaires. Travel behavior questions were the same as those used in the statewide Caltrans study, but supplemented by questions related to residential location decisions and perceptions of the new rail station access.See Jonathan Levine, et al., Land Use and Transportation Alternatives: Constraint or Expansion of Household Choice, Mineta Transportation Institute (MTI) Report 01-19. San Jose, CA: San Jose State University, 2002. See Appendix A for the survey instrument.
The following are the primary research questions of the resident surveys:
Does residential location near the Gold Line increase the probability of transit use in work and nonwork trips, as compared to transit-mode share in the surrounding city? To what degree does location in a TOD bring about a change in travel behavior (versus relocation of prior transit users)?
What is the influence of other factors, such as the service qualities of competing modes, demographic characteristics, policy features, and neighborhood attributes?
What factors influence residential location in a TOD? How did travel behavior differ from the previous residence to the current one?
Surveys were sent to all residential units within each building. The questionnaire included a cover letter describing the project and giving instructions for completing the survey; a letter of endorsement from the mayor of Pasadena for Pasadena surveys or the councilman for the city of Los Angeles for the Los Angeles surveys; and a prepaid, preaddressed envelope. The first mailing was sent in early June, and follow-up survey packages were sent to all nonrespondents three weeks later.
Two additional strategies were used to increase response rates. First, everyone who returned a completed questionnaire was eligible for a raffle drawing with two $500 cash prizes. Second, surveys were translated into Spanish and there was a notice at the top of the introductory letter informing potential respondents (in Spanish) of how to obtain a Spanish-language survey. The survey team received only three requests for Spanish surveys; none of these was returned. The resulting overall response rate is 13.9 percent. The survey instrument is provided in Appendix A.
Key
Participant Interviews
Inquiries into the development and management responses to new rail access were conducted through oral interviews with project developers and property managers operating within Gold Line station areas. The following are the primary research questions for the developer and property manager surveys:
How does the Gold Line influence the market for TOD in the entire corridor?
How did proximity to the Gold line affect a specific TOD project, considering factors such as planning, design, development economics, financing, development entitlements, and marketing?
We selected developers that represent
the range of development conditions along the Gold Line. Neighborhoods vary
from residential districts with a small-town feeling to urban-scale mixed-use
areas. We generated our list of interview targets by talking with planners who
worked at LACMTA, the city of
An interview guide was developed to ensure that consistent questions were asked of respondents. The questions were open-ended, allowing the respondent to explain the situation as he or she sees it. (See Appendix B for a copy of the interview guide.) The primary objectives of the developer interviews were to gain insights into how light-rail transit station proximity influenced development practices and to discover planning issues that support or impede the continuation of TOD in the corridor. For each developer, Web research was completed on project characteristics and a telephone interview was conducted with a senior manager directly involved in the development. Each interview took about 30 minutes.
The Gold Line and Its Station Areas
The Pasadena Gold Line extends from
Union Station in downtown
Source: Metropolitan Transportation Authority (http://www.mta.net)
Map of Pasadena
Among the best-known TODs in the corridor is
The following overviews illustrate the wide range of population, housing, and development characteristics along the Gold Line corridor. These overviews are based on data from the 2000 U.S. Census (see Table 2 on page 25) and site observations, and are presented here from west to east.See U.S. Census data for the one-third-mile radii around station areas were compiled using the LandView 6 program produced by the U.S. Census Bureau. Land View 6 estimates population characteristics within a radius "by tallying Census 2000 data for those block [group] points whose coordinates fall within the circle defined by the radius." Data are derived from Summary Files 1 and 3 of the 2000 U.S. Census. The surveyed TOD sites are described briefly.
Southwest Museum
The
Most of the housing stock consists of one-unit buildings and structures that were built prior to 1960. Of the occupied housing units, 58 percent are rental, and nearly 60 percent of residents had moved into their current residence within the past five years (as of 2000). The overall densities are 10,918 persons per square mile and 3,850 housing units per square mile. Immediately next to the station is a day-care facility in a historic structure that was developed to take advantage of Gold Line proximity.
Three TOD sites or buildings were included in this study: two
along
Evaluations of the pedestrian
environment between each of the sites and the rail station reveal that routes
are relatively safe from traffic, with sidewalks on both sides of the street
for the entire route and crosswalks and signals at every intersection, but they
do not rate well in terms of comfort, aesthetics, or sense of safety. None of
the block faces has more than one street tree or street light; there are no
benches or other furniture; landscaping is sporadic; and graffiti, trash, and
bars on windows were common.
Highland Park
The housing stock is quite diverse in
terms of the number of units per building (38 percent are single-unit, 41
percent have 2 to 19 units, and 21 percent have 20 units or more) and the age
of the structures (18 percent were built between 1980 and 1994, 27 percent
between 1960 and 1979, and 55 percent before 1960). As of 2000, the station
area was not characterized by new development: No structures were built between
1995 and 2000. Eighty percent of the occupied housing units are rentals, and
nearly 60 percent of the population moved into their current residence within
the past five years (from 1995 to 2000). Portions of the station area are under
a historic overlay zone and a TOD ordinance. The station is one block from the
busy main street of the commercial district,
Unfortunately, household addresses were
successfully obtained for only one (15-unit) building in this station area.
Even with a 13 percent response rate, this provided only two surveys. These
surveys are clustered with
The pedestrian environment surrounding
the
Mission
The Mission station area is located in
the old town area of
Nearly one-half of the housing stock consists of one-unit dwellings, and two-thirds of the housing structures were built prior to 1960. Sixty-three percent of occupied housing units were rentals as of 2000, and 61 percent of all residents (including renters and owners) moved to their current residence between 1995 and 2000. The community is in the path of a planned extension of the 710 Freeway, but has participated with others opposed to the freeway in preventing its construction. Some in the community saw the Gold Line as an alternative to freeway building.
Ten TOD sites or buildings were
surveyed within the
The station has no public parking yet and only one connecting bus
route. However, LACMTA has teamed up with a TOD developer to fund public
station parking in a TOD next to the station, which is under construction. The
two
Fillmore
Primarily comprising medical and hospital services, Fillmore's station area has the lowest population density as of 2000 (3,279 persons per square mile). City planners saw this corridor as being appropriate for biotechnology research and development, but that development has been slow in occurring. The station area also has a low share of young children (just 4 percent of the total population); the highest educational attainment among the studied station areas, with 38 percent of the adult population earning a bachelor's degree or higher; and a high share of single-person households (58 percent). Fillmore's racial and ethnic make-up is quite diverse, with Asian persons representing 15 percent of the population, Black or African American persons comprising 12 percent, and Hispanic or Latino comprising 19 percent. Fillmore has the highest share of single-occupancy-vehicle (SOV) commuters (87 percent) and the lowest share of carpool commuters (1 percent). Only 5 percent of the station-area population commuted by transit in 2000.
Fillmore has few total housing units (664) compared to the other nine station areas. One-third of these units are located in large multi-unit buildings (20 units or more), and just 22 percent are single-unit dwellings. Fillmore also has a high share of renters (80 percent) and newer residents (70 percent moved in between 1995 and 2000). The Fillmore station has 160 public parking spaces and five connecting bus routes.
Four TODs were surveyed in the Fillmore Station. All sites are located close to one another, along South Marengo Avenue, about 0.2 mile southeast of the station. Two were built in the early 1960s, two in the early 1980s, and one is a condominium complex. Sizes range from 12 to 30 units per building. The pedestrian routes have average to excellent landscaping (ranging from 3 to 5 on the subjective 5-point ranking), and there was no evidence of trash, graffiti, vacant buildings, or barred windows. Although there is little retail nearby and only one street tree and street light per block face, it was a fairly pleasant route overall.
Del
Del Mar is similar to Fillmore in several ways: low population density (4,388 persons per square mile), few children under 5 years of age (4 percent of the population), high educational attainment (29 percent have a bachelor's degree or higher), a large share of single-person households (62 percent), and a relatively diverse ethnic composition (10 percent Black or African American, 11 percent Asian, 28 percent Hispanic). It also has a large majority of workers (80 percent) who commute by SOV and few (2 percent) who carpool. However, this station area is going through substantial change, with construction of the Del Mar Station TOD surrounding the station. The completed project will include public parking for the Gold Line and the restoration of the Santa Fe Depot train station.
The size of
Four large buildings (80 to 150 units)
were surveyed in the
Memorial
Park
The Memorial Park station area is located at the northern edge of
historic Old Pasadena. It has a low population density and few young children,
but has a higher share of seniors than Fillmore and
The most interesting aspect of the Memorial Park station area, however, is that 29 percent of its working population walks to work. This is not surprising given the transit- and walking-dependent population (Memorial Park has more low-income households, elderly persons, and households with no vehicles than any other station area), but this behavior also appears to be supported by Memorial Park's central location in Pasadena's mixed-use, pedestrian-friendly environment.
The large majority of housing units (82
percent) near the Memorial Park station are located in buildings with 20 or
more units. As of 2000, this station area had the largest share of recent
development, with 24 percent of its housing structures constructed between 1995
and 2000. This percentage drops to just 5 percent in the station areas with the
next largest shares of recent development (Fillmore and
Only two sites were surveyed within the
Memorial Park station area. One of these (
With a major TOD built directly above the station and several TODs located within one-quarter-mile east of the station, the pedestrian environment in and around the Memorial Park station was one of the strongest of all the TODs studied. Landscaping was excellent, with nearly every block face along each of the routes ranking either a 4 or 5, and there was no evidence of trash, graffiti, vacant buildings, or barred windows. Street trees and lights were not over-abundant (only one street tree and light per block face on average), but intersection wait times were rarely more than 30 seconds.
Lake
The Lake station area has much higher
densities than the three
Building sizes are fairly well distributed across single-unit dwellings, small multi-unit buildings, and large buildings (20 or more units). The area had received some recent development as of 2000, but nearly all the housing structures were built before 1995: 24 percent between 1980 and 1995, 30 percent between 1960 and 1979, and 42 percent before 1960. As with nearly all the studied station areas, the large majority of residents (85 percent) are renters, and more than half moved into their current residence between 1995 and 2000.
These demographics have to be understood in terms of the role that the 134 Freeway plays in bifurcating the station area. The area to the south of the station is high-density office and new residential units, while the area to the north of the station is a more traditional residential neighborhood with a lower income profile. The station itself in located in the freeway median, below street level.
Four TODs were surveyed within the
Residents living to the north or south
have a substantial walking distance to reach the station. As with the rest of
the stations to the east, this station has a harsher station platform
environment, related to noise and safety issues associated with being in a
freeway median. As the most westerly of the three stations located along
Interstate 210, conditions along the walking routes probably varied the most of
any of the stations examined, with landscaping ranging from 1 to 5 and
intersection wait times ranging from no wait time to 60 seconds. Newer sites in
this area, including
Allen
The Allen station area has the
second-highest population density (after
The majority of the housing stock (as
of 2000) was built before 1960, with just 3 percent built from 1995 to 2000.
The station area has a higher share of single-family dwellings (63 percent of
all housing units), and a lower share of renters (57 percent) relative to most
other station areas. As in other station areas, the majority of Allen's
residents had moved in within the last 5 years (as of 2000). The Allen station
has the most residential neighborhood feel of all the stations. South of the
station, a large cluster of automobile repair shops are interspersed with the
residential development.
The three TODs surveyed in the Allen station area are all located along North Allen and all are condominiums. Two of the sites were built in the early 1970s and consist of 75 units each; the third was built in 1989 and has just 10 units. The Allen station has no parking and four connecting bus routes.
Walking south along
Sierra
Madre Villa
In spite of their physical separation
at opposite ends of the Gold Line, the residents and housing stock around
Sierra Madre Villa are quite similar to that of the
Sierra Madre Villa's housing stock is almost entirely composed of older, single-family dwellings: 88 percent of the housing units are one-unit (attached or detached) buildings, and more than three-quarters of the housing stock was constructed before 1960. This is the only station area with more owners than renters (60 percent and 40 percent respectively), and the only one where more than half of all residents had lived in their place of residence for more than 5 years as of 2000. This housing stock coexists with major employment centers and a big box retail outlet.
Six multi-unit TODs were identified and
surveyed in the Sierra Madre Villa station area: one on
Because of the lack of multi-unit
housing within walking distance of the Sierra Madre Villa station, the selected
sites are all 0.6 to 0.7 mile from the station--by far the longest walking
distance of any residential buildings that we examined. The five sites on
Overview of Residential Surveys
This chapter summarizes the results of the residential surveys conducted at TODs in the Gold Line corridor.
Response Rates
Surveys were distributed to 1,595 multifamily housing units located within one-third mile of Gold Line stations. Each survey respondent was asked to provide detailed information on up to three "main" trips for a predetermined day of the week. (In order to collect as much information as possible on commute travel, and to remain consistent with the previous statewide studies, weekend days were not included.) A trip is defined in the survey instrument as "one direction of travel (for instance, from home to work or from work to the grocery store or from a restaurant to home)." Respondents determine which trips constitute a "main" trip, but are asked to include at least one direction of their travel to work or school if they made a work or school trip during their recorded day of travel.
A total of 221 surveys were
successfully completed and returned, for an overall response rate of 13.9
percent. Response rates were highest (30.1 percent) in the
The 14 percent response rate achieved is slightly higher than that of the 2003 statewide TOD study. Caution must be exercised in drawing interpretations from individual stations because of low numbers of responses. The station groupings shown in Table 3 are the greatest degree to which results can be disaggregated.